Publication type
Report
Authors
Publication date
July 15, 2021
Summary:
HESA commissioned the Institute for Social and Economic Research (ISER) at the University of Essex to carry out research to ascertain whether application of statistical weighting to the Graduate Outcomes survey data can effectively mitigate the consequences of non-response. Data from the second year of the survey, relating to graduates from the 2018/19 academic year, were used in the research. The findings were subsequently replicated on data from the first year of the survey (2017/18 graduates) to establish the robustness of the conclusions. A rich set of auxiliary variables were available for weighting and weights were developed based on a number of different non-response models. The weights were tested to see whether they improved the accuracy of estimates of the proportion of graduates in employment and/or study or in highly skilled employment and/or study. It was found that weighting – based on any of the alternative models – improved accuracy in only a minority of cases, and that the magnitude of the reduction when it occurred was very small. It was consequently concluded that with these data there is no need to use weighted estimation: the accuracy of estimates does not substantially differ between weighted and unweighted estimates. Users of Graduate Outcomes Survey data may be reassured by these conclusions, as they indicate that there is no evidence of substantial non-response bias in the survey data.
Subjects
Link
https://www.hesa.ac.uk/files/ISER-Graduate-Outcomes-weighting-report-20210720.pdf
#568058